INDEX

Explanations

specific references to academic or scientific data and metrics

oai_token-act-pair · gpt-4o-mini Triggered by @bot

foreign language references

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

setVerticalGroup

-0.38

melidir

-0.35

 parvenir

-0.32

 conclusión

-0.30

 grandeza

-0.30

 emplois

-0.29

 olabilir

-0.29

 müm

-0.29

 pourraient

-0.28

POSITIVE LOGITS

 myſelf

0.83

 Verſ

0.77

 itſelf

0.76

 Monfieur

0.73

 ſei

0.73

ſelf

0.73

niſſe

0.71

 ſich

0.70

 ſind

0.70

 ſta

0.68

Activations Density 0.110%

specific references to academic or scientific data and metrics

foreign language references

No Comments

No Known Activations